# common/utils.py
import urllib.parse
import re
import constants

def normalize_url(url: str) -> str:
    parsed = urllib.parse.urlparse(url)
    clean = parsed.scheme + "://" + parsed.netloc + parsed.path
    return clean.rstrip("/")

def extract_real_video_url(url: str) -> str:
    if not url: return ""
    parsed_url = urllib.parse.urlparse(url)
    query_params = urllib.parse.parse_qs(parsed_url.query)
    for param_name, param_values in query_params.items():
        for value in param_values:
            if value.startswith(('http://', 'https://')):
                decoded_value = urllib.parse.unquote(value)
                if any(ext in decoded_value.lower() for ext in constants.VIDEO_EXTENSIONS):
                    return decoded_value
    decoded_url = urllib.parse.unquote(url)
    if decoded_url != url and any(ext in decoded_url.lower() for ext in constants.VIDEO_EXTENSIONS):
        return decoded_url
    return url

# ⚖️ Elite Fix: Deterministic Credibility Engine (Rules before AI)
def get_domain_credibility(url):
    domain = urllib.parse.urlparse(url).netloc.lower()
    if domain.endswith(('.gov', '.edu', '.mil', 'arxiv.org', 'nature.com', 'science.org', 'nasa.gov')):
        return "High_Scientific_Gov"
    elif domain.endswith(('.org', '.com', '.net')) and any(x in domain for x in ['news', 'bbc', 'cnn', 'reuters', 'natgeo', 'space.com']):
        return "Medium_Mainstream_Media"
    elif any(x in domain for x in ['reddit', 'medium', 'quora', 'blogspot', 'wordpress', 'pinterest']):
        return "Low_Fringe_UGC"
    else:
        return "Unknown_Evaluate"

# 🛑 NEW: Bouncer for Banned Domains (TikTok, Pinterest, Dailymotion etc.)
def is_banned_domain(url: str) -> bool:
    try:
        domain = urllib.parse.urlparse(url).netloc.lower()
        # इन साइट्स पर कोई आर्टिकल नहीं होता, सिर्फ UI टेक्स्ट होता है
        banned_keywords = ['tiktok.com', 'pinterest.', 'dailymotion.com', 'instagram.com', 'facebook.com', 'wattpad.com']
        return any(banned in domain for banned in banned_keywords)
    except:
        return False

# 🎥 NEW: YouTube ID Extractor
def get_youtube_video_id(url: str):
    try:
        parsed_url = urllib.parse.urlparse(url)
        if 'youtube.com' in parsed_url.netloc:
            if parsed_url.path == '/watch':
                return urllib.parse.parse_qs(parsed_url.query).get('v', [None])[0]
            elif parsed_url.path.startswith(('/embed/', '/v/')):
                return parsed_url.path.split('/')[2]
        elif 'youtu.be' in parsed_url.netloc:
            return parsed_url.path[1:]
    except:
        pass
    return None